ABSTRACT
BACKGROUND: Three trials of tight glucose control (TGC) found clinical benefit of normalization of blood glucose levels in the intensive care unit (ICU). Implementation of TGC was imperfect in subsequent trials, since attained blood glucose levels (BGLs) never reached the targets as in the original trials of TGC. We investigated whether implementation of the TGC guideline as used in the original trials of TGC is feasible and safe. METHODS: In this study 3 ICUs adopted and implemented the TGC guideline as used in the original trials of TGC using a multifaceted practice change strategy; 3 ICUs that did not change their blood glucose control guideline served as controls. TGC was practiced by physicians and nurses during the first 12-month (period-2), thereafter exclusively by nurses (period-3). Blood glucose metrics 12-month before (period-1) and 24-month after implementation of the guideline were compared. RESULTS: The analysis included 1321 in period-1, 1169 and 1006 patients in period-2, and -3, respectively, in the intervention ICUs, and 3110 patients in the control ICUs. After implementation of the new TGC guideline, patients in intervention ICUs had lower median BGLs (105 [IQR: 85-130] mg/dL vs. 119 [99-150] mg/dL in period-1, P<0.001; and vs. 113 [95-141] mg/dL in control ICUs, P<0.001). The incidence of severe hypoglycemia initially increased, but again decreased when exclusively nurses practiced TGC, and was not associated with increased mortality or morbidity. CONCLUSIONS: Implementation of the original TGC guideline is feasible and safe. Our study suggests a learning effect over time.
Subject(s)
Blood Glucose/analysis , Critical Care/standards , Critical Illness , Guideline Adherence , Hyperglycemia/prevention & control , Practice Guidelines as Topic , Aged , Aged, 80 and over , Circadian Rhythm , Critical Care/methods , Critical Illness/mortality , Critical Illness/nursing , Female , Hospital Mortality , Hospitals, Community , Humans , Hyperglycemia/drug therapy , Hyperglycemia/epidemiology , Hypoglycemia/epidemiology , Hypoglycemia/prevention & control , Insulin/administration & dosage , Insulin/adverse effects , Insulin/therapeutic use , Intensive Care Units , Male , Middle Aged , Netherlands/epidemiologyABSTRACT
BACKGROUND: To study current clinical practice in blood glucose (BG) control in adult intensive care units (ICUs) in the Netherlands. METHODS: We performed a national survey focusing on blood glucose targets, insulin administration, BG control guidelines, and opinions regarding BG control aiming for normoglycaemia (known as intensive insulin therapy, IIT). RESULTS: The completed questionnaire was returned by 88/113 (78%) of the participating centres. In 98% (86/88) of the ICUs some sort of BG control was being practised. Half of the ICUs (42/86, 48%) used tight BG targets as with IIT; 28/86 (33%) and 13/86 (15%) used more liberal targets of 4.4 to 7.0 mmol/l and 4.4 to 8.0 mmol/l, respectively. Eighty-two (93%) reported having a local guideline on BG control (or IIT). The BG threshold to start insulin was 7.0+/-1.3 mmol/l vs 7.8+/-1.3 mmol/l in ICUs that practised IIT vs ICUs that practised less tight BG control, respectively (p=0.005). In 28/86 (33%) measurement of the BG values was done according to a strict time schedule (i.e., BG values were measured on predefined time points). While respondents were fairly agreed on the benefits of IIT, opinions regarding ease of implementation and time needed to apply this strategy varied. In addition, severe hypoglycaemia was considered a serious side effect of IIT. CONCLUSION: Approximately half of the ICUs in the Netherlands reported having implemented IIT. However, the full guideline as used in the original studies on IIT was hardly ever implemented. Concerns about severe hypoglycaemia, at least in part, hampers implementation of IIT.
